Commit 32f20dee authored by Eva Zangerle's avatar Eva Zangerle
Browse files

reformatting of notebook 08

parent 22a91f61
......@@ -3,7 +3,9 @@
{
"cell_type": "markdown",
"id": "edd718da-1295-49c4-b556-3cc7b718f93c",
"metadata": {},
"metadata": {
"tags": []
},
"source": [
"# Hypothesis Testing\n",
"Lecture Data Engineering and Analytics<br>\n",
......@@ -19,11 +21,11 @@
"source": [
"# import required packages\n",
"import os\n",
"import statistics\n",
"\n",
"import matplotlib.pyplot as plt\n",
"import numpy as np\n",
"import pandas as pd\n",
"import seaborn as sns\n",
"from scipy import stats\n",
"from sklearn.utils import resample"
]
},
......@@ -119,18 +121,6 @@
"id": "f65ee2f1-c5bb-4146-9645-711fa3a2dce6",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" income type\n",
"19678 48300.0 Data\n",
"5695 40000.0 Data\n",
"13497 60000.0 Data\n",
"6298 30000.0 Data\n",
"1687 75000.0 Data\n"
]
},
{
"data": {
"text/html": [
......@@ -158,28 +148,28 @@
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>19678</th>\n",
" <td>48300.00</td>\n",
" <th>37771</th>\n",
" <td>62000.00</td>\n",
" <td>Data</td>\n",
" </tr>\n",
" <tr>\n",
" <th>5695</th>\n",
" <td>40000.00</td>\n",
" <th>23161</th>\n",
" <td>136000.00</td>\n",
" <td>Data</td>\n",
" </tr>\n",
" <tr>\n",
" <th>13497</th>\n",
" <td>60000.00</td>\n",
" <th>5603</th>\n",
" <td>100000.00</td>\n",
" <td>Data</td>\n",
" </tr>\n",
" <tr>\n",
" <th>6298</th>\n",
" <td>30000.00</td>\n",
" <th>17459</th>\n",
" <td>40000.00</td>\n",
" <td>Data</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1687</th>\n",
" <td>75000.00</td>\n",
" <th>37843</th>\n",
" <td>79000.00</td>\n",
" <td>Data</td>\n",
" </tr>\n",
" <tr>\n",
......@@ -189,27 +179,27 @@
" </tr>\n",
" <tr>\n",
" <th>4995</th>\n",
" <td>78028.85</td>\n",
" <td>63650.00</td>\n",
" <td>Mean of 20, 5000 samples</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4996</th>\n",
" <td>82030.60</td>\n",
" <td>82726.15</td>\n",
" <td>Mean of 20, 5000 samples</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4997</th>\n",
" <td>75391.50</td>\n",
" <td>58868.15</td>\n",
" <td>Mean of 20, 5000 samples</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4998</th>\n",
" <td>56003.35</td>\n",
" <td>67831.25</td>\n",
" <td>Mean of 20, 5000 samples</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4999</th>\n",
" <td>65848.20</td>\n",
" <td>71026.10</td>\n",
" <td>Mean of 20, 5000 samples</td>\n",
" </tr>\n",
" </tbody>\n",
......@@ -218,18 +208,18 @@
"</div>"
],
"text/plain": [
" income type\n",
"19678 48300.00 Data\n",
"5695 40000.00 Data\n",
"13497 60000.00 Data\n",
"6298 30000.00 Data\n",
"1687 75000.00 Data\n",
"... ... ...\n",
"4995 78028.85 Mean of 20, 5000 samples\n",
"4996 82030.60 Mean of 20, 5000 samples\n",
"4997 75391.50 Mean of 20, 5000 samples\n",
"4998 56003.35 Mean of 20, 5000 samples\n",
"4999 65848.20 Mean of 20, 5000 samples\n",
" income type\n",
"37771 62000.00 Data\n",
"23161 136000.00 Data\n",
"5603 100000.00 Data\n",
"17459 40000.00 Data\n",
"37843 79000.00 Data\n",
"... ... ...\n",
"4995 63650.00 Mean of 20, 5000 samples\n",
"4996 82726.15 Mean of 20, 5000 samples\n",
"4997 58868.15 Mean of 20, 5000 samples\n",
"4998 67831.25 Mean of 20, 5000 samples\n",
"4999 71026.10 Mean of 20, 5000 samples\n",
"\n",
"[9000 rows x 2 columns]"
]
......@@ -279,7 +269,7 @@
" sample_mean_20_2,\n",
" ]\n",
")\n",
"print(results.head())\n",
"\n",
"results"
]
},
......@@ -299,7 +289,7 @@
"outputs": [
{
"data": {
"image/png": "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\n",
"image/png": "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\n",
"text/plain": [
"<Figure size 288x720 with 5 Axes>"
]
......@@ -344,7 +334,10 @@
"outputs": [],
"source": [
"def bootstrap(data, function, no_draws):\n",
" \"\"\"function draw no_draws samples of data, applies func and stores result in array\"\"\"\n",
" \"\"\"\n",
" function draw no_draws samples of data, applies func and\n",
" stores result in array\n",
" \"\"\"\n",
" results = []\n",
" for nrepeat in range(no_draws):\n",
" sample = resample(data, replace=True)\n",
......@@ -372,7 +365,7 @@
},
{
"data": {
"image/png": "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\n",
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
......@@ -384,8 +377,9 @@
}
],
"source": [
"# also: we can use this example to showcase the central limit theorem (again), changing the number of draws\n",
"mean_distribution = bootstrap(loans_income, np.mean, 200)\n",
"# also: we can use this example to showcase the central limit theorem (again),\n",
"# by changing the number of draws\n",
"mean_distribution = bootstrap(loans_income, np.mean, 2000)\n",
"mean_distribution.plot.hist()"
]
},
......@@ -401,7 +395,7 @@
"text": [
"Bootstrap Statistics:\n",
"original data: 68760.51844\n",
"std. error: 144.6409686847048\n"
"std. error: 148.64731550006957\n"
]
}
],
......@@ -437,8 +431,29 @@
{
"data": {
"text/plain": [
"0.05 68515.351217\n",
"0.95 68966.071996\n",
"0 68606.30468\n",
"1 68851.19516\n",
"2 68927.11040\n",
"3 68870.23618\n",
"4 68718.73022\n",
" ... \n",
"195 69094.20692\n",
"196 68896.40868\n",
"197 68484.23478\n",
"198 68710.79954\n",
"199 68646.20612\n",
"Length: 200, dtype: float64"
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"text/plain": [
"0.05 68525.858346\n",
"0.95 69019.373161\n",
"dtype: float64"
]
},
......@@ -451,6 +466,7 @@
"print(loans_income.mean())\n",
"# create a sample of 20 loan income data\n",
"mean_distribution = bootstrap(loans_income, np.mean, 200)\n",
"mean_distribution\n",
"mean_distribution.quantile([0.05, 0.95])\n",
"confidence_interval = list(mean_distribution.quantile([0.05, 0.95]))"
]
......@@ -463,7 +479,7 @@
"outputs": [
{
"data": {
"image/png": "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\n",
"image/png": "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\n",
"text/plain": [
"<Figure size 720x576 with 1 Axes>"
]
......@@ -509,1131 +525,6 @@
"\n",
"plt.tight_layout();"
]
},
{
"cell_type": "code",
"execution_count": 13,
"id": "2b647865-1f2c-484c-9fcb-6453e8a50355",
"metadata": {},
"outputs": [],
"source": [
"group1 = [19, 20, 14, 23, 15, 18]\n",
"group2 = [13, 14, 8, 17, 9, 12]"
]
},
{
"cell_type": "code",
"execution_count": 14,
"id": "f6118595-e8bc-4607-bc1a-a4612d8a1f02",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"10.966666666666667"
]
},
"execution_count": 14,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"text/plain": [
"10.966666666666667"
]
},
"execution_count": 14,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"m1 = np.mean(group1)\n",
"m2 = np.mean(group2)\n",
"var1 = statistics.variance(group1, xbar=None)\n",
"var2 = statistics.variance(group2, xbar=None)\n",
"var1\n",
"var2"
]
},
{
"cell_type": "code",
"execution_count": 15,
"id": "f6f41623-08f8-4d9a-bbc6-22e0b6928320",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"3.138156196894831"
]
},
"execution_count": 15,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"(m1 - m2) / np.sqrt((var1 / 6) + (var2 / 6))"
]
},
{
"cell_type": "code",
"execution_count": 16,
"id": "226ed025-c9d2-40a7-b4cf-f6e029f338c9",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Ttest_indResult(statistic=3.138156196894831, pvalue=0.010543184275035807)"
]
},
"execution_count": 16,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"stats.ttest_ind(group1, group2)"
]
},
{
"cell_type": "code",
"execution_count": 17,
"id": "8df460fc-829c-4592-ad72-80b5f43d4dbf",
"metadata": {},
"outputs": [],
"source": [
"def read_derm_data():\n",
" np.random.seed(1)\n",
"\n",
" # Histopathological Attributes: (values 0, 1, 2, 3)\n",
" # Clinical Attributes: (values 0, 1, 2, 3, unless indicated)\n",
" features = [\n",
" \"erythema\",\n",
" \"scaling\",\n",
" \"definite borders\",\n",
" \"itching\",\n",
" \"koebner phenomenon\",\n",
" \"polygonal papules\",\n",
" \"follicular papules\",\n",
" \"oral mucosal involvement\",\n",
" \"knee and elbow involvement\",\n",
" \"scalp involvement\",\n",
" \"family history\", # 0 or 1\n",
" \"melanin incontinence\",\n",
" \"eosinophils in the infiltrate\",\n",
" \"PNL infiltrate\",\n",
" \"fibrosis of the papillary dermis\",\n",
" \"exocytosis\",\n",
" \"acanthosis\",\n",
" \"hyperkeratosis\",\n",
" \"parakeratosis\",\n",
" \"clubbing of the rete ridges\",\n",
" \"elongation of the rete ridges\",\n",
" \"thinning of the suprapapillary epidermis\",\n",
" \"spongiform pustule\",\n",
" \"munro microabcess\",\n",
" \"focal hypergranulosis\",\n",
" \"disappearance of the granular layer\",\n",
" \"vacuolisation and damage of basal layer\",\n",
" \"spongiosis\",\n",
" \"saw-tooth appearance of retes\",\n",
" \"follicular horn plug\",\n",
" \"perifollicular parakeratosis\",\n",
" \"inflammatory monoluclear inflitrate\",\n",
" \"band-like infiltrate\",\n",
" \"Age\", # linear; missing marked '?'\n",
" \"TARGET\", # See mapping\n",
" ]\n",
"\n",
" targets = {\n",
" 1: \"psoriasis\", # 112 instances\n",
" 2: \"seboreic dermatitis\", # 61\n",
" 3: \"lichen planus\", # 72\n",
" 4: \"pityriasis rosea\", # 49\n",
" 5: \"cronic dermatitis\", # 52\n",
" 6: \"pityriasis rubra pilaris\", # 20\n",
" }\n",
"\n",
" data = os.path.join(data_dir, \"dermatology.data\")\n",
" df = pd.read_csv(data, header=None, names=features, na_values=[\"?\"])\n",
" df[\"TARGET\"] = df.TARGET.map(targets)\n",
"\n",
" derm = df.copy()\n",
" derm.loc[derm.Age == \"?\", \"Age\"] = None\n",
" derm[\"Age\"] = derm.Age.astype(float)\n",
" return derm"
]
},
{
"cell_type": "code",
"execution_count": 18,
"id": "ef10fdef-711f-4546-a9bd-d54a31366cec",
"metadata": {},
"outputs": [],
"source": [
"derm = read_derm_data()"
]
},
{
"cell_type": "code",
"execution_count": 19,
"id": "90fcb5e3-4e4c-466a-8f76-c9481ed04593",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>erythema</th>\n",
" <th>scaling</th>\n",
" <th>definite borders</th>\n",
" <th>itching</th>\n",
" <th>Age</th>\n",
" <th>TARGET</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>247</th>\n",
" <td>2</td>\n",
" <td>2</td>\n",
" <td>2</td>\n",
" <td>0</td>\n",
" <td>62.0</td>\n",
" <td>psoriasis</td>\n",
" </tr>\n",
" <tr>\n",
" <th>127</th>\n",
" <td>2</td>\n",
" <td>2</td>\n",
" <td>2</td>\n",
" <td>2</td>\n",
" <td>44.0</td>\n",
" <td>lichen planus</td>\n",
" </tr>\n",
" <tr>\n",
" <th>230</th>\n",
" <td>3</td>\n",
" <td>2</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>30.0</td>\n",
" <td>seboreic dermatitis</td>\n",
" </tr>\n",
" <tr>\n",
" <th>162</th>\n",
" <td>3</td>\n",
" <td>2</td>\n",
" <td>2</td>\n",
" <td>2</td>\n",
" <td>22.0</td>\n",
" <td>lichen planus</td>\n",
" </tr>\n",
" <tr>\n",
" <th>159</th>\n",
" <td>3</td>\n",
" <td>2</td>\n",
" <td>2</td>\n",
" <td>1</td>\n",
" <td>47.0</td>\n",
" <td>seboreic dermatitis</td>\n",
" </tr>\n",
" <tr>\n",
" <th>296</th>\n",
" <td>2</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>3</td>\n",
" <td>19.0</td>\n",
" <td>cronic dermatitis</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" erythema scaling definite borders itching Age TARGET\n",
"247 2 2 2 0 62.0 psoriasis\n",
"127 2 2 2 2 44.0 lichen planus\n",
"230 3 2 0 1 30.0 seboreic dermatitis\n",
"162 3 2 2 2 22.0 lichen planus\n",
"159 3 2 2 1 47.0 seboreic dermatitis\n",
"296 2 1 1 3 19.0 cronic dermatitis"
]
},
"execution_count": 19,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# use iloc on both axes and sample\n",
"derm.iloc[:, [0, 1, 2, 3, -2, -1]].sample(6)"
]
},
{
"cell_type": "code",
"execution_count": 20,
"id": "43734f3f-a997-436c-93c9-04785c6a5c47",
"metadata": {},
"outputs": [],
"source": [
"clean, suspicious = [], {}\n",
"for col in derm.columns:\n",
" values = derm[col].unique()\n",
" if set(values) <= {0, 1, 2, 3}:\n",
" clean.append(col)\n",
" else:\n",
" suspicious[col] = values"
]
},
{
"cell_type": "code",
"execution_count": 21,
"id": "46b6f7d4-77a0-4e7d-b70e-039caa83568d",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"No problem detected:\n"
]
},
{
"ename": "NameError",
"evalue": "name 'pprint' is not defined",
"output_type": "error",
"traceback": [
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
"\u001b[0;31mNameError\u001b[0m Traceback (most recent call last)",
"\u001b[0;32m/tmp/ipykernel_128943/136470058.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m 1\u001b[0m \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"No problem detected:\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m \u001b[0mpprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mclean\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;36m8\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 3\u001b[0m \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34mf\"... {len(clean) - 8} other fields\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;31mNameError\u001b[0m: name 'pprint' is not defined"
]
}
],
"source": [
"print(\"No problem detected:\")\n",
"pprint(clean[:8])\n",
"print(f\"... {len(clean) - 8} other fields\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "d1180b53-e9b0-407d-bd1a-22474288e195",
"metadata": {},
"outputs":